• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Records Before and After cleaning

OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
It was a busy week but I did the tests that I wanted to last night and I'll show my results in this post and the one below. I didn't have too many dirty records laying around so I went to the shop and got a few $2 records in ok shape so I can do a few more samples.
I find a lot of what I did to be mildly conclusive at best but I'm hoping for some critique and feedback. This is what I did.
1.). Record 1min 30s of my first Sample Record from the second track in the middle of the record. I preferred the 2nd track instead of the first because it being closer to the null points. I can't really tell you I can hear a sound change from any of the tracks but I know that properly setup the middle of a side should theoretically should have lower distortion. Please tell me if this is dumb. I also took a smaller sample because if I'm able to get the Deltawave null test to work I assume shorter tracks are easier to match.

2.) I cleaned the record on my Kladio. I set the clean cycle to 5 minutes. I have other ways to clean records, this is just my preferred way. I believe it does a good job (I have not tested yet, that's what I'm trying to do here) subjectively. Also, I love the ease of use. My loricraft is more work to clean although it works great in combo. I just wanted to keep it simple for this experiment. So 5 min clean and 5 min dry.

3.) I recorded the same track. I trim both tracks and best I could to match.

4.) I look at the statistics and compared.

5.) I declicked in Isotope RX. I'm going to also declick in Click Repair.

6.) I put them into Deltawave and compared the difference.

All this data will be in other posts.

The record I used was Earthy by the Prestige All Stars. This copy has not been cleaned to my knowledge but it's from 1957 and I am not the original owner as I wasn't born yet. It had visual evidence of dust and is what I would call a VG+ but maybe I'm too generous.
The problem I overlooked until later is that this is a mono record which I kinda forgot. I didn't use my mono cartridge though, I used a stereo cartridge.

My chain of gear went like this.
Lyra Kleos -> Technics 1210G Table and Arm -> Manley Steelhead -> Lynx Hilo -> Mac Mini -> Amadeus

I have other options but this is my pretty standard chain for digitizing. I do not have available a non RIAA phono preamp but am going to get a WaxWing sooner or later.
I'm going to use mostly stereo records from now on.
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
Here is the Dirty, and basic statistics.

Sample_1_Dirty_Waveform.png



Sample_1_Dirty_Statistics.png
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
Here is the click removal. I was thinking it would have had a bigger difference. It appears he difference was about a 10% improvement in clicks after cleaning.

Dirty Clicks
Sample_1_Dirty_Clicks.jpg


Cleaned Clicks

Sample_1_Cleaned_Clicks.jpg
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
And I did Deltawave but I don't know how well it went. How good was the nulling? @pkane had advised it should work but I can't tell if it did and how well. Below is the Spectrum Delta. Deltawave does so much that I don't know what's interesting to screen shot and show. I'm having issue with interpreting this and would love so critique.

Delta_Spectrum.jpg
 

egellings

Major Contributor
Joined
Feb 6, 2020
Messages
4,076
Likes
3,320
I'm curious if anyone has though about an experiment that studied the benefits of cleaning records. What would you measure? Distortion before and after? Noise relative between the two records?
Samples, it would be tough for me to come up with a perfect sampling since my collection is not representative of a more broad collection. Also, you can't really control the dirtiness of the records so your sample would need to be large. What section of the record would you record to see the noise?
I'm curious at this point but was wonder if anyone had opinions or examples of this? I can't find anything that resembles a broad study on the value (or lack thereof) of record cleaning.
I don't know how you would measure that without a test record expressly made for the purpose. I guess just listening would be a good way to judge that. If the record's noisy or quiet, you'll immediately know it. If nothing else, cleaning records would prolong the life of the records & stylus, a worthwhile goal, even if noise audibility is low.
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
I don't know how you would measure that without a test record expressly made for the purpose. I guess just listening would be a good way to judge that. If the record's noisy or quiet, you'll immediately know it. If nothing else, cleaning records would prolong the life of the records & stylus, a worthwhile goal, even if noise audibility is low.
I totally agree with you I just am trying to show it. so every record would be different in how effective the cleaning is. But I have questions like How much improvement is say used records As an average percentage. Does that distribution have too much variance to be useful. Does cleaning benefit new records and a rough expected value on how much it helps. Also, click repair does t measure surface noise improve,ent. Can that be measured? I dont know how so I’m just trying stuff.
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,706
Likes
10,396
Location
North-East
And I did Deltawave but I don't know how well it went. How good was the nulling? @pkane had advised it should work but I can't tell if it did and how well. Below is the Spectrum Delta. Deltawave does so much that I don't know what's interesting to screen shot and show. I'm having issue with interpreting this and would love so critique.

View attachment 345549

There are a lot of plots and measurements in DeltaWave. A lot depends on what you're trying to determine. This looks like a comparison of a cleaned up (pop removal?) capture from an LP vs. dirty? Maybe share the two files, and we can help you with where to look.
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,706
Likes
10,396
Location
North-East
Thank you for your offer. I can't tell you what I'm looking for exactly. I'm looking for evidence that the surface noise is lower on the Cleaned file vs the Dirty file. Or not.

https://drive.google.com/drive/folders/1r4LbRVaiWodb1oMQH6rvqa3pVp8J1_hr?usp=sharing
Thanks for uploading the files!

There does appear to be a significant difference between the two captures. Which one is more accurate is hard to tell without a reference to compare it to, of course. It may also help to have a few captures of each, before and after cleaning, so that differences can be confirmed not to be due to variations in the TT playback.

The difference in frequency shows that the two diverge mostly towards higher frequencies:

1706492814395.png


The "Dirty" recording (blue) has slightly lower amplitude above 40kHz or so:
1706498005940.png


You can see the frequency differences over time in the spectrogram:
1706497780210.png


Audibility-weighted PK Metric shows a few places where there are significant (audible) differences, but overall, the RMS PK Metric shows only a slightly audible difference of -40dB:
1706497854765.png
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
There does appear to be a significant difference between the two captures. Which one is more accurate is hard to tell without a reference to compare it to, of course. It may also help to have a few captures of each, before and after cleaning, so that differences can be confirmed not to be due to variations in the TT playback.
thanks for taking a look. I have a few more samples and will start doing multiple captures for each. I’m aiming for 10 samples and going to evaluate how predictive this endeavor is.
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
Below is a copy of my spreadsheet to keep track of this project. I've only done 7. The problem is time and I don't always have tons of dirty records laying around. I'm happy to go shopping for some $1 bin (now $2 bin at my spot) and will probably splurge when I go into the city this weekend. But I can only bring home so many records before my wife kills me. She is not a fan of the quantity of my collection.
Look at sample 5, that one suprised me a lot. It did sound good for something from like 1954 after I cleaned it but that's a huge click reduction.
I also just eyeballed a huge click on Sample 7 that was reduced 6 decibels but still there after cleaning.
I'm not doing new records (I clean new records) because it's more controversial so I'm keeping it separate. If anyone wants to see the Deltawave analysis on any of these let me know.

Is this interesting to anyone. I can keep doing this if people are learning getting something from it. I think of course my sample has to be much larger to see any patter besides cleaning records helps somewhat to a different degree per record.
Also, I don't want to change methodology. I like this one because it's easiest. The Neil Antin method is too much for me to do a ton of them like that and document it. It is unquestionably better, but more complicated and time consuming. I use other methods (I have a Loricraft too) but I want to do everyone the same in this experiment.


Sample #ArtistAlbumDateConditionPre Wash ClicksPost Wash ClicksPercent DifferenceNotes
1Prestige All StarsEarthy1/24/2024VG+201111810010.00%Original, Mono
2Dave LiebmannLookout Farm1/25/2024VG+10402817921.37%
3Leo KotkeLeo Kotke1/25/2024VG+6452530817.73%
4Frank ZappaStairway to Heaven1/25/2024M-4416328525.61%12" Single, Live so lots of clicks do to applause. Maybe bad example
5Clifford Brown, Art Farmer and Swedish All StarsCliff Brown Art Farmer And The Swedish All Stars1/26/2024VG+11822140588.12%10", Mono
6Joseph Jarman, Famoudou Don MoyeEgwu-Anwu (Sun Song) D11/27/2024VG+6288507319.32%2 Discs, Live
7Joseph Jarman, Famoudou Don MoyeEgwu-Anwu (Sun Song) D21/28/2024VG+11531105488.52%2 Discs, Live
 

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,116
Likes
1,410
Below is a copy of my spreadsheet to keep track of this project. I've only done 7. The problem is time and I don't always have tons of dirty records laying around. I'm happy to go shopping for some $1 bin (now $2 bin at my spot) and will probably splurge when I go into the city this weekend. But I can only bring home so many records before my wife kills me. She is not a fan of the quantity of my collection.
Look at sample 5, that one suprised me a lot. It did sound good for something from like 1954 after I cleaned it but that's a huge click reduction.
I also just eyeballed a huge click on Sample 7 that was reduced 6 decibels but still there after cleaning.
I'm not doing new records (I clean new records) because it's more controversial so I'm keeping it separate. If anyone wants to see the Deltawave analysis on any of these let me know.

Is this interesting to anyone. I can keep doing this if people are learning getting something from it. I think of course my sample has to be much larger to see any patter besides cleaning records helps somewhat to a different degree per record.
Also, I don't want to change methodology. I like this one because it's easiest. The Neil Antin method is too much for me to do a ton of them like that and document it. It is unquestionably better, but more complicated and time consuming. I use other methods (I have a Loricraft too) but I want to do everyone the same in this experiment.


Sample #ArtistAlbumDateConditionPre Wash ClicksPost Wash ClicksPercent DifferenceNotes
1Prestige All StarsEarthy1/24/2024VG+201111810010.00%Original, Mono
2Dave LiebmannLookout Farm1/25/2024VG+10402817921.37%
3Leo KotkeLeo Kotke1/25/2024VG+6452530817.73%
4Frank ZappaStairway to Heaven1/25/2024M-4416328525.61%12" Single, Live so lots of clicks do to applause. Maybe bad example
5Clifford Brown, Art Farmer and Swedish All StarsCliff Brown Art Farmer And The Swedish All Stars1/26/2024VG+11822140588.12%10", Mono
6Joseph Jarman, Famoudou Don MoyeEgwu-Anwu (Sun Song) D11/27/2024VG+6288507319.32%2 Discs, Live
7Joseph Jarman, Famoudou Don MoyeEgwu-Anwu (Sun Song) D21/28/2024VG+11531105488.52%2 Discs, Live
If you are confident in your skills you can offer cleaning as a service. I'm sure there are locals who would take you up on it.
 
OP
mfalcon

mfalcon

Member
Joined
Jan 17, 2024
Messages
46
Likes
17
If you are confident in your skills you can offer cleaning as a service. I'm sure there are locals who would take you up on it.
I'm probably better off just going to work but it sounds like a good retirement gig. I told my daughter she's welcome to use my RCM's to make some money if she wants. Who knows.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,197
Likes
3,767
In the few times I've played a record in the last decade, it was to digitize it, and I always bought an unopened copy on discogs, and did the wood glue treatment to both sides before recording.

So I have no way to compare, i.e., I'm running on superstition. I can say this digitized audio required less pop/click fixing compared to much older digitizations I did with traditionally 'cleaned' records.....using an entirely different TT setup, so again, not anywhere near science.
 

GPJ

Member
Joined
Jan 3, 2019
Messages
47
Likes
102
The Parks Puffin (and the Waxwing) have a tool for grading LP quality. I ran across a test that an ordinary user did after performing ultrasonic cleaning of various LP's and doing a before and after using the Puffin. I captured a screenshot (attached pic) of the results (I thinks it's under the Parks twitter site, it's been quite some time ago). Anyway, I use an ultrasonic cleaner and do like the results I get.
 

Attachments

  • ultrasonic.jpg
    ultrasonic.jpg
    495.2 KB · Views: 33
Last edited:

Curvature

Major Contributor
Joined
May 20, 2022
Messages
1,116
Likes
1,410
The Parks Puffin (and probably the Waxwing) have a tool for grading LP quality. I ran across a test that an ordinary user did after performing ultrasonic cleaning of various LP's and doing a before and after using the Puffin. I captured a screenshot (attached pic) of the results (I thinks it's under the Parks twitter site, it's been quite some time ago). Anyway, I use an ultrasonic cleaner and do like the results I get.
Nice hardware version of Excel you got there.

It's appropriate, seeing that vinyls are the hardware version of mp3s.
 
  • Like
Reactions: GPJ
Top Bottom